Laser powder bed melting thermal scanning chromatography quality evaluation method and system

Abstract

The invention discloses a laser powder bed melting thermal scanning chromatography quality evaluation method and system. According to the laser powder bed fusion thermal scanning chromatography quality evaluation method, obtained layer section radiation data is fused and updated in a gradient image mode by using a signal processing technology and an image fusion technology, a layer section fused gradient image is generated, gain noise reduction processing is performed on the layer section fused gradient image, and the layer section fused gradient image is obtained. The method comprises the following steps: acquiring a layer cross section fusion gradient image, converting the layer cross section fusion gradient image into a pseudo-color image, performing gradient anomaly region extraction operation on the pseudo-color image, establishing a correlation model among gradient anomaly, printing anomaly and printing defects, and realizing online quality evaluation by utilizing the established correlation model and combining layer cross section radiation data acquired in real time. Gradient abnormity in the printing process is monitored in real time, and possible printing defects are predicted, so that measures can be taken in time for intervention and adjustment, the defective rate of products is reduced, and cost reduction and efficiency improvement are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of additive manufacturing, and in particular to a laser powder bed melting thermal scanning tomography quality evaluation method and system. Background Art

[0002] Laser powder bed fusion (LPBF) is an additive manufacturing technology that uses a laser beam to melt and solidify metal powder layer by layer at an extremely high speed to form the required parts. It uses a wide range of materials, integrates design and manufacturing, has obvious advantages in the refined manufacturing of complex structures, and has been widely used in the manufacturing process of complex components of various types of equipment.

[0003] However, the thermal physics and non-equilibrium metallurgical processes of materials are very complex during the LPBF process. There are complex interactions between the laser and metal powder, the molten pool and the substrate. The heat conduction of the powder and the heat transfer process inside the molten pool are extremely complex. When the laser energy density reaches a certain threshold, the molten pool fluctuates violently under the coupling of recoil pressure, surface tension and Marangoni effect, which is easy to form keyhole pore defects. In addition, the abnormal thermal history caused by rapid heating and cooling may also produce other defects such as unfused holes and microcracks. Due to the possibility of these internal quality defects, non-destructive testing such as X-rays or industrial CT is often required before delivery of aviation parts. However, industrial CT has many limitations in testing costs, testing efficiency and accessibility of complex structures. At the same time, this post-testing solution can only be inspected after forming, and it is impossible to find and optimize problems in time during the forming process. Even if problems are found, the high manufacturing cost cannot be recovered.

[0004] Therefore, there is an urgent need to develop a laser powder bed melting thermal scanning tomography quality evaluation method, which aims to promptly detect possible printing defects during the forming process, so as to take timely measures to intervene and adjust, thereby achieving cost reduction and efficiency improvement. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a laser powder bed melting thermal scanning tomography quality evaluation method and system.

[0006] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0007] A laser powder bed fusion thermal scanning tomography quality evaluation method comprises the following steps: acquiring layer section radiation data in a layer section processing process in real time; fusing and updating the acquired layer section radiation data in a gradient image manner by using signal processing technology and image fusion technology to generate a layer section fusion gradient image; performing gain noise reduction processing on the layer section fusion gradient image and converting the layer section fusion gradient image into a pseudo color image; performing a gradient abnormality region extraction operation on the pseudo color image and establishing a correlation model between gradient abnormality and printing abnormality and printing defects, thereby realizing online quality evaluation.

[0008] A laser powder bed melting thermal scanning tomography quality evaluation system comprises: a tomography detection sensor, which is installed on the upper top plate of a forming cavity and can perform photosensitivity measurement on the entire forming cross section, and is used to collect layer cross section radiation data during the layer cross section processing; a control module, which is used to obtain the layer cross section radiation data in real time; using signal processing technology and image fusion technology to fuse and update the acquired layer cross section radiation data in the form of gradient images to generate a layer cross section fusion gradient image; performing gain noise reduction processing on the layer cross section fusion gradient image, and converting the layer cross section fusion gradient image into a pseudo color image; performing gradient abnormality area extraction operation on the pseudo color image, and establishing a correlation model between gradient abnormality and printing abnormality and printing defects, thereby realizing online quality evaluation.

[0009] The beneficial technical effect of the present invention is that the above-mentioned laser powder bed melting thermal scanning tomography quality evaluation method uses signal processing technology and image fusion technology to fuse and update the acquired layer section radiation data, generate a layer section fusion gradient image, and convert the layer section fusion gradient image after gain noise reduction processing into a pseudo-color image, perform gradient abnormality area extraction operation on the pseudo-color image, establish a correlation model between gradient abnormality and printing abnormality and printing defects, and use the established correlation model in combination with the real-time acquired layer section radiation data to achieve online quality evaluation, real-time monitoring of gradient abnormalities in the printing process, and prediction of possible printing defects, so that timely measures can be taken to intervene and adjust, reduce the defective rate of products, and thus achieve cost reduction and efficiency improvement. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 It is a schematic structural diagram of the laser powder bed melting thermal scanning tomography quality evaluation system of the present invention;

[0011] Figure 2 It is a schematic flow chart of the laser powder bed melting thermal scanning tomography quality evaluation method of the present invention;

[0012] Figure 3 Schematic diagram of tomographic fusion.

[0013] Description of reference numerals:

[0014] 1-laser, 2-laser beam, 3-scanning device, 4-part forming surface, 5-forming platform, 6-layer cross-section radiation, 7-observation window, 8-optical component, 9-tomography detection sensor, 10-top plate of forming cavity. DETAILED DESCRIPTION

[0015] In order to enable those skilled in the art to more clearly understand the objectives, technical solutions and advantages of the present invention, the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0016] The invention provides a laser powder bed melting thermal scanning tomography quality evaluation system.

[0017] like Figure 1 As shown, in one embodiment of the present invention, the laser powder bed melting thermal scanning tomography quality evaluation system includes a tomography detection sensor 9 and a control module (not shown in the figure), the tomography detection sensor 9 is installed on the upper top plate 10 of the forming cavity, and the tomography detection sensor 9 can perform photosensitivity measurement on the entire forming cross-section, and is used to collect layer cross-section radiation data during the layer cross-section processing process; the control module is used to obtain the layer cross-section radiation data in real time, and use signal processing technology and image fusion technology to fuse and update the acquired layer cross-section radiation data in the form of gradient images to generate a layer cross-section fusion gradient image, perform gain noise reduction processing on the layer cross-section fusion gradient image, and convert the layer cross-section fusion gradient image into a pseudo-color image, perform gradient abnormality area extraction operation on the pseudo-color image, and establish a correlation model between gradient abnormalities and printing abnormalities and printing defects, thereby realizing online quality evaluation.

[0018] The tomography detection sensor 9 uses any one or any combination of a CMOS camera, a photodiode array, a near-infrared thermal imager or an infrared thermal imager. Since the emission spectra of different materials will be different when the high-energy beam heats and melts them, different sensors can be selected as the tomography detection sensor 9 according to the spectral emission curves of different metal materials during the melting and solidification process: first measure and analyze the emission spectrum of the high-energy beam melted metal to determine the spectral emission curve of the metal material during the melting and solidification process; then determine the target band based on the spectral emission curve of the metal material during the melting and solidification process; finally select a suitable sensor as the tomography detection sensor 9 according to the target band.

[0019] Among them, a band with higher spectral intensity can be selected as the target band from the spectral emission curve of the metal material during the melting and solidification process. According to different processing material objects, the target band can be a plasma band, a laser reflection band, a near-infrared band, an infrared band, etc.; it can also be a narrowband band or a wide range band.

[0020] The target band is the band of the layer cross-sectional radiation 6 to be collected by the layer-to-layer detection sensor 9, and the layer-to-layer detection sensor 9 should be able to collect all layer cross-sectional radiation 6 within the target band. Therefore, after determining the target band, a suitable sensor can be selected as the layer-to-layer detection sensor 9 according to the target band. If the target band is within the visible light and near-infrared bands, the layer-to-layer detection sensor 9 can use a CMOS camera, a photodiode array, or a near-infrared thermal imager; if the target band exceeds the near-infrared band, the layer-to-layer detection sensor 9 uses an infrared thermal imager, etc.

[0021] For example Figure 1 As shown, an observation window 7 is provided on the upper top plate 10 of the forming cavity, and an optical component 8 is provided at the front end of the tomography detection sensor 9. The optical component 8 is used to transmit the layer cross-section radiation 6 of a specific band to the tomography detection sensor 9.

[0022] The wavelength band that the observation window 7 can transmit is consistent with the wavelength band of the layer cross-section radiation 6. When the wavelength band of the layer cross-section radiation 6 is within the plasma band, the laser reflection band and the near-infrared band, the observation window 7 can use an optical window made of fused quartz, N-BK7 and the like; when the wavelength band of the layer cross-section radiation 6 includes a part greater than 2000nm, an optical window made of germanium, zinc sulfide, zinc selenide and the like can be used.

[0023] The optical component 8 is used to determine the wavelength band of the layer cross-section radiation 6 transmitted to the layer cross-section detection sensor 9, so as to transmit the layer cross-section radiation 6 of a specific wavelength band to the layer cross-section detection sensor 9. The optical component 8 determines the wavelength band of light that can pass through the optical component 8 by adjusting the internal filter lens combination. The wavelength band of light that can pass through the optical component 8 can be a narrow band (such as 800nm±25nm, 950nm±25nm, etc.) or a wide range band (such as 1200nm-1800nm, 1200nm-2600nm, 2000nm-5000nm, etc.).

[0024] During the laser powder bed melting process, the laser 1 emits a laser beam 2, and the metal powder is heated and melted on the part forming surface 4 through a scanning device 3 through a certain scanning strategy. During the heating, melting and cooling and solidification process, layer cross-section radiation 6 will be formed on the layer cross-section. The layer cross-section radiation 6 is transmitted to the observation window 7 located on the upper top plate 10 of the forming cavity through spatial light, and then transmitted to the tomography detection sensor 9 through the optical component 8, and collected by the tomography detection sensor 9 to monitor the gradient change of the layer cross-section radiation. It should be noted that the wavelength band of the laser beam 2 emitted by the laser 1 is a high-energy beam laser of 1064nm or other wavelength bands. The wavelength band of the layer cross-section radiation 6 can be a plasma band, a laser reflection band, a near-infrared band, an infrared band, etc.

[0025] The invention provides a laser powder bed melting thermal scanning tomography quality evaluation method.

[0026] like Figure 2 As shown, in one embodiment of the present invention, the laser powder bed melting thermal scanning tomography quality evaluation method includes steps S10 to S40:

[0027] S10. Acquire the layer cross-section radiation data in the layer cross-section processing process in real time.

[0028] In specific implementation, use Figure 1 The tomography detection sensor 9 in the laser powder bed melting thermal scanning tomography quality evaluation system in the shown embodiment collects the layer section radiation data during the layer section processing in real time, and then communicates with the tomography detection sensor 9 to obtain the layer section radiation data during the layer section processing in real time.

[0029] S20, using signal processing technology and image fusion technology to fuse and update the acquired layer cross-section radiation data in a gradient image manner to generate a layer cross-section fused gradient image.

[0030] like Figure 3 As shown, from left to right, the image frames acquired within the set unit time are fused and refreshed as units, and a layer cross-section fusion gradient image is gradually formed. The set unit time can be a short period (any value from 1 to 1000 ms), a long period (1 s to 60 s), or an ultra-long period (more than 1 min, or even the sintering time of a complete layer).

[0031] In the image fusion process, if integral fusion is adopted, the integral processing is continuously performed on each pixel in the frame image during the fusion process, and the generated fusion image is an integral distribution image. If extreme value fusion is adopted, the extreme value is continuously sought for each pixel during the fusion process, and the generated fusion image is an extreme value distribution image. If weighted fusion is adopted, the weighted coefficient is multiplied for each pixel during the integral fusion and extreme value fusion process to generate a weighted integral image and a weighted extreme value image. If normalized distribution processing is adopted for the image during the fusion process, the image gradient distribution is normalized to between 0 and 1, and a normalized fusion image is generated. When the layer section processing is completed, the generated layer section fusion gradient image is any one or any combination of integral distribution image, extreme value distribution image, weighted integral image, weighted extreme value image, and normalized fusion image.

[0032] S30, performing gain noise reduction processing on the layer-section fusion gradient image, and converting the layer-section fusion gradient image into a pseudo-color image.

[0033] The gain denoising method such as median filtering, Gaussian filtering, wavelet transform or deep learning denoising is used to perform gain denoising on the layer section fusion gradient image to improve the contrast of the effective radiation information in the image while ensuring that the gain algorithm does not introduce new noise or distortion. At the same time, the layer section fusion gradient image is converted into a pseudo-color image for analysis.

[0034] S40, performing a gradient abnormality region extraction operation on the pseudo-color image, establishing a correlation model between gradient abnormality and printing abnormality and printing defects, and thus achieving online quality evaluation.

[0035] Image algorithms such as pixel value threshold, region segmentation, morphological operations, or machine learning algorithms such as convolutional neural networks are used to extract gradient abnormality areas from pseudo-color images, and a correlation model between gradient abnormalities and printing abnormalities and printing defects is established to achieve online quality evaluation.

[0036] The pixel value threshold processing method is to stipulate that the overall grayscale distribution or gradient distribution of the image is in a threshold interval. If it exceeds the interval, the exceeding area is a gradient abnormality area. Pixel value threshold is a basic image segmentation method. Its basic principle is to divide the pixels in the image into different categories (such as target and background) by setting one or more grayscale thresholds. In specific operations, each pixel in the image will be traversed and classified according to the size relationship between its grayscale value and the threshold. If the grayscale value of the pixel is greater than the threshold, it is classified as the target class; otherwise, it is classified as the background class.

[0037] The region segmentation method divides a set of pixels corresponding to an object or an object surface from an image. There are various methods for region segmentation, including contour fitting, region growing, split-merge, etc. These methods usually divide regions based on the grayscale of the fused image. For example, the region growing method starts with a set of seed points and gradually "grows" the region according to the similarity between pixels (such as grayscale similarity) until the set abnormal region stop condition (such as area, shape, etc.) is met.

[0038] The morphological operation is a shape and structure-based image processing method, the core of which is the definition and operation of structural elements (SE). Morphological operations include dilation, erosion, opening and closing operations. Dilation can enlarge objects in the image to make them more connected; erosion, on the contrary, can shrink objects to make them more refined. Opening and closing operations are combined operations of dilation and erosion, which are used to remove noise and smooth edges and fill holes respectively. Morphological operations are widely used in image processing, especially in image segmentation, edge detection, noise removal, etc., and are mainly used for further accurate processing of extracted abnormal areas.

[0039] Convolutional neural network (CNN) extracts and classifies features of input images through structures such as convolutional layers, pooling layers, and fully connected layers. In image processing, CNN can automatically learn the hierarchical feature representation in images, from low-level edge and texture features to high-level shape and structural features. Through training, CNN can learn the mapping relationship between images and specific tasks (gradient anomaly detection), thereby accurately extracting gradient anomaly areas in images.

[0040] After the gradient abnormality area is extracted, the extracted gradient abnormality area is associated with sample data such as abnormal phenomena in the actual printing process or the internal quality or physical and chemical properties of the printed part, and the relationship between the gradient abnormality and the printing quality is mined through machine learning algorithms (such as support vector machines, random forests, neural networks, etc.), so as to establish a correlation model between the gradient abnormality and printing abnormalities and printing defects. After the correlation model is established, the correlation model can be used to achieve online quality evaluation, real-time monitoring of gradient abnormalities in the printing process, and prediction of possible printing defects, so that timely measures can be taken for intervention and adjustment.

[0041] The laser powder bed melting thermal scanning tomography quality evaluation method of the present invention utilizes signal processing technology and image fusion technology to fuse and update the acquired layer section radiation data, generate a layer section fusion gradient image, and convert the layer section fusion gradient image after gain noise reduction processing into a pseudo-color image, perform gradient abnormality area extraction operation on the pseudo-color image, establish a correlation model between gradient abnormality and printing abnormality and printing defects, and utilize the established correlation model in combination with the layer section radiation data acquired in real time to realize online quality evaluation, real-time monitoring of gradient abnormality in the printing process, and prediction of possible printing defects, so that timely measures can be taken to intervene and adjust, reduce the defective rate of products, and thus achieve cost reduction and efficiency improvement.

[0042] The above description is only a preferred embodiment of the present invention, and does not limit the present invention in any form. Those skilled in the art can make various equivalent changes and improvements based on the above embodiments, and all equivalent changes or modifications made within the scope of the claims should fall within the protection scope of the present invention.

Claims

1. A laser powder bed melting thermal scanning tomography quality evaluation method, characterized in that: The steps include: S10, acquiring the layer cross-section radiation data in the layer cross-section processing process in real time; S20, using signal processing technology and image fusion technology to fuse and update the acquired layer cross-section radiation data in a gradient image manner to generate a layer cross-section fusion gradient image; S30, performing gain noise reduction processing on the layer-section fusion gradient image, and converting the layer-section fusion gradient image into a pseudo-color image; S40, performing a gradient abnormality region extraction operation on the pseudo-color image, establishing a correlation model between gradient abnormality and printing abnormality and printing defects, and thus achieving online quality evaluation.

2. The laser powder bed melting thermal scanning tomography quality evaluation method according to claim 1, characterized in that: The step S20 further includes: fusing and refreshing the image frames acquired within a set unit time, and gradually forming a layer cross-section fused gradient image.

3. The laser powder bed melting thermal scanning tomography quality evaluation method according to claim 2, characterized in that: Integral fusion is used in the fusion process to continuously perform integration processing on each pixel in the frame image, and the generated layer section fusion gradient image is an integral distribution image.

4. The laser powder bed melting thermal scanning tomography quality evaluation method according to claim 2, characterized in that: In the fusion process, extreme value fusion is adopted to continuously find the extreme value pixel by pixel, and the generated layer section fusion gradient image is an extreme value distribution image.

5. The laser powder bed melting thermal scanning tomography quality evaluation method according to claim 2, characterized in that: In the fusion process, the image is processed by normalized distribution, and the image gradient distribution is normalized to between 0 and 1. The generated layer section fusion gradient image is the normalized fusion image.

6. The laser powder bed melting thermal scanning tomography quality evaluation method according to claim 1, characterized in that: In step S30, a median filtering method, a Gaussian filtering method, a wavelet transform method or a deep learning denoising method is used to perform gain denoising processing on the layer section fused gradient image.

7. The laser powder bed melting thermal scanning tomography quality evaluation method according to claim 1, characterized in that: In step S40, a pixel value threshold algorithm, a region segmentation algorithm, a morphological operation algorithm or a convolutional neural network algorithm is used to perform a gradient abnormality region extraction operation on the pseudo-color image.

8. A laser powder bed melting thermal scanning tomography quality evaluation system, characterized in that: Included are: A layer detection sensor is installed on the top plate of the forming cavity. The layer detection sensor can perform photosensitive measurement on the entire forming cross section and is used to collect layer cross section radiation data during the layer cross section processing; The control module is used to acquire the layer section radiation data in real time; use signal processing technology and image fusion technology to fuse and update the acquired layer section radiation data in the form of gradient images to generate a layer section fusion gradient image; perform gain noise reduction processing on the layer section fusion gradient image, and convert the layer section fusion gradient image into a pseudo-color image; perform gradient abnormality area extraction operation on the pseudo-color image, establish a correlation model between gradient abnormality and printing abnormality and printing defects, and then realize online quality evaluation.

9. The laser powder bed fusion thermal scanning tomography quality evaluation system according to claim 8, characterized in that: An optical component is provided at the front end of the tomographic detection sensor, and the optical component is used to transmit layer cross-section radiation of a specific waveband to the tomographic detection sensor.

10. The laser powder bed fusion thermal scanning tomography quality evaluation system according to claim 8, characterized in that: The tomography detection sensor adopts any one or any combination of a CMOS camera, a photodiode array, a near-infrared thermal imager or an infrared thermal imager.

Images